Benchmarking with real data or copies of real data is useful but it becomes problematic if you want to compare the performance of one site with another site and can't use the same data set at both sites. Having a consistent data set that is sanitized and can be used at any site can be a useful for performance benchmarks and consistent testing.
Below is a download link for a LinearHash table and accompanying source data table of a 1 million row person information table. Records are sequentially keyed from 1 to 1,000,000.
Linear Hash LK/OV Files, Dictionary, and RAW sources download link:
CSDataSample_PersonInfo.7z - 146m
Sample data was generated using generatedata.com and then imported into an OpenInsight Linear Hash table.